How to Write an RFP That Gets Good Responses

Ask a procurement team why an RFP produced disappointing bids and you will usually hear that the market is weak.

Occasionally that is true. More often the document asked the wrong questions and allowed too little time. It pushed unreasonable risk onto the supplier, and gave no clue how bids would be judged. Strong suppliers read that and quietly decline.

The UK Government's Sourcing Playbook puts the underlying point in one sentence: "Suppliers have a choice whether to work with us or not." That applies to private buyers just as much.

#First, is an RFP the right tool?

RFI — request for information

Use when you do not yet know who supplies this or how it is normally done. It is research, not a competition, and should be short.

RFQ — request for quotation

Use when you know exactly what you want and the only real variable is price and delivery. A defined specification, a simple comparison.

RFP — request for proposal

Use when you know the outcome you need but not the best way to achieve it, and you want suppliers to propose an approach. Price is one factor among several.

Running an RFP when you needed an RFQ wastes everyone's time. Running an RFQ when you needed an RFP buys you the cheapest version of the wrong solution.

#The work that happens before you write

A clear specification

The Playbook is blunt about why this matters. A clear specification has to tell bidders enough to decide whether they want to bid at all. Without that shared understanding, it says, you cannot relate the price offered to your own understanding of costs. And if you cannot do that, you are always at risk of missing the outcome you want at the price you need.

Write the specification in terms of the outcome you need wherever you can. Specify inputs only where the input genuinely matters, because every input you fix removes an option a supplier might have offered you.

An internal view of what it should cost

The Playbook's approach is a "should cost model" — an independent estimate of the right cost or cost range, and of what belongs in the whole-life cost. Without one, you cannot tell an efficient bid from a bid that is underpriced and will fail.

This does not need to be sophisticated. A defensible estimate built from labour, materials, volumes and margin is enough to tell you when a number does not make sense.

Talking to the market first

Early engagement is not cheating. The Playbook recommends testing outline evaluation criteria with potential bidders as part of early market engagement, and refining the evaluation model iteratively over time.

Suppliers will tell you which requirements are expensive, which are unusual, and which make the bid unattractive. Far better to hear it before you publish.

#What goes in the RFP

#Evaluation criteria and scoring

This is the part most RFPs do worst, and it is the part that decides the outcome.

The Playbook describes what a sound approach looks like: objective criteria relevant to the requirement, weightings applied according to the importance of each criterion, and "a scoring approach that promotes effective differentiation".

That last phrase is the one to take seriously. If every bidder scores 4 out of 5 on everything, quality has no influence. The decision then defaults to price. That is exactly the bias toward low cost bids that good evaluation design is meant to avoid.

Making scores mean something

1. Write a definition for each score before you see any bid. State what a 2 looks like, what a 4 looks like and what has to be present for a 5.
2. Weight criteria by real importance, not by how easy they are to measure. If continuity matters more than price, say so in the weightings.
3. Score each question independently and only then look at the totals.
4. Have at least two evaluators score separately, then meet to agree a moderated score with a written reason.
5. Never average raw scores across evaluators without discussing the gaps. A wide spread usually means the question was ambiguous.
6. Keep the record. If you cannot explain a decision three months later, you cannot debrief a losing bidder or defend a challenge.

Publishing the weightings in the RFP is not giving something away. It tells serious bidders where to spend their effort, which improves what you receive.

#Timelines: the cost you cannot see

Suppliers need enough time to develop and price a solution, raise clarifications and respond properly. The Playbook notes that inadequate timescales produce rushed solutions that may miss opportunities to innovate or to deliver something better.

It is harder still on changes made mid-process. Changing timelines once a procurement is under way incurs great costs for suppliers. It damages the buyer's credibility and discourages bidding. And because delays add cost to suppliers, they raise prices unnecessarily.

Practical minimums for a mid-sized commercial RFP:

If you cannot meet your own published dates, tell bidders as soon as you know. Silence after a submission deadline is the single most common complaint suppliers have about buyers.

#Keep the cost of bidding down

Every question you add has a cost that somebody pays, and eventually that somebody is you.

The cost of bidding for public sector contracts is frequently cited as a reason for not bidding and as a barrier to entry for SMEs and VCSEs... By making our procurement processes unnecessarily complicated or protracted, we risk minimising the pool of bidders and stifling competition.

UK Cabinet Office, The Sourcing Playbook

Three tests for every question in your response template. Will the answer change the decision? Could we already find this out ourselves? Would we notice if nobody answered it?

Questions that fail all three are the bulk of most RFP question sets. Company history, generic methodology statements and requests to "describe your commitment to quality" produce identical marketing text from every bidder and separate nobody.

Ask instead about the specific thing you are buying. What would you do in the first 90 days on this account? Who exactly will do the work and what else are they assigned to? What has gone wrong on a similar contract and what changed afterwards?

#Risk allocation is a pricing decision

The Playbook is direct about this. Risks should sit with the party best able to manage them. Inappropriate allocation of risk, it says, remains one of the main concerns of suppliers looking to do business with the public sector.

Transferring a risk the supplier cannot control does not remove the risk. It buys you a price increase and a supplier who will argue when the risk materialises.

Where risk is genuinely shared, say so in the RFP and share the risk register. You will get more accurate prices and fewer surprises at contract stage.

#Performance measures belong in the RFP

Bidders price what they are measured on, so the measures must be in the document they are pricing against — not invented afterwards.

Keep the set small. The Playbook warns that having too many measures, which it puts at more than 10 to 15 per service, "will lead to overcomplicated contracts and ambiguity with suppliers".

#The mistakes that show up most often

• No evaluation method published, so bidders guess what matters and answer the wrong things.
• A specification copied from the incumbent's current contract, which guarantees the incumbent wins and nobody innovates.
• Terms issued after shortlisting, so the real negotiation happens when you have no alternative left.
• A question set built by adding every stakeholder's request and removing none.
• A deadline set by an internal date nobody explained, giving bidders two weeks for a complex scope.
• No debrief for losing bidders, which is the cheapest way to make good suppliers skip your next RFP.

#After the decision

Tell everyone the outcome, including the unsuccessful bidders, and do it promptly. Offer a short debrief explaining where their bid scored well and where it did not, in terms of your published criteria.

This costs perhaps twenty minutes per bidder. It is the main reason a good supplier bids again next time, and it forces your own evaluation to be defensible.